For a complex trait, the pedigree-based llinkage analysis is effective for localizing disease-related genes to within 1cM. Fine-mapping of the disease mutation within this region is often achieved by linkage disequilibrium analysis using haplotypes. Haplotypes can also be very useful for stratifying patients and directing drugs to appropriate subpopulation. The discovery of human single nucleotides polymorphism (SNP) as genetic marker's greatly increases the density of the genetic map and enhances our ability in developing novel disease diagnostic tools and individualized drug therapies. However, with 1.65 million human SNPs single nucleotides polymorphism (SNP) markers identified, the haplotype determination and analysis present a daunting challenge to scientists. The general goal of this research is to advance our computational capability for both the construction of population haplotypes and the utilization of the whole haplotype information in LD mapping. The specific aims of the proposed research are (a) to propose more efficient and scalable computational strategies for haplotype determination, (b) to demonstrate that in silico approaches to haplotype determinations are effective and economical alternatives to experimental ones, (c) to establish a computational platform that allows for a direct output of haplotypes from high-throughput genotyping readouts, and (d) to design a Bayesian framework for simultaneous haplotype determination, LD mapping, and patients clustering.